There is a new regulatory framework governing hospital transfusion laboratory practice which was implemented after the publication of two European Union Directives: 2002/98/EC and 2004/33/EC.⁶ In the UK these are the Blood Safety and Quality Regulations 2005 (Statutory Instruments 2005/50, 2005/1098 and 2006/2013) and they set standards for quality and safety of human blood and blood components in hospital ‘blood banks’ as well as ‘blood establishments’ (the UK Blood Services).⁷ ‘Blood banks’ are regulated in their roles of storing, distributing and performing compatibility tests, on blood and blood components for use in hospitals. The implications for hospital transfusion practice are the requirement for ‘vein to vein’ traceability of blood components, the importance of maintaining the ‘cold chain’ for all therapeutic blood components, the need to store transfusion records for 30 years and the requirement for a quality management system.^6,⁷ UK laboratories have to assess themselves and submit an annual compliance report to the competent authority, which is the Medicines and Healthcare Products Regulatory Agency (MHRA).⁸ The MHRA carries out laboratory inspections to assess compliance with these regulations.

The UK government, through the National Blood Transfusion Committee, continues to promote the safe and effective transfusion of blood components and has published three Health Service Circulars.⁹ These documents aim to ensure that a team approach to blood transfusion safety is taken via the local clinical governance arrangements and to promote good transfusion practice via Hospital Transfusion Committees and Hospital Transfusion Teams supported by Regional Transfusion Committees.

Technology and automation in blood transfusion laboratories

Important changes have taken place in the blood transfusion laboratory in the last 10–15 years, as outlined in the previous edition of this book. As a result transfusion laboratory practice is safer, largely as a result of the implementation of new technologies for testing, automated systems to replace manual systems and the widespread use of information technology (IT) systems to support transfusion laboratory practice.

Barcoded labels on blood components, reagents, patient samples and equipment are now commonplace and these result in safer transfer of information, free from the transcription errors associated with manual methods.

Column agglutination (CAT) and solid-phase technology can both be used on automated machines and CAT can also be used by manual techniques. In UK hospital transfusion laboratories these technologies have replaced tube techniques and liquid-phase microplates for antibody screening and crossmatching (Figs 22.2, 22.3).¹⁰

Figure 22.2 Change in indirect antiglobulin test (IAT) technology for antibody screening. Data from UK NEQAS surveys 1998–2009.¹⁰ These surveys of UK practice show that technology used for antibody screening and serological crossmatching is now predominantly column agglutination (CAT) with very few laboratories using tube techniques. LPM, liquid-phase microplate.

Figure 22.3 Change in technology for serological compatibility testing. Data from UK NEQAS surveys 1998–2009.¹⁰ These surveys of UK practice show that technology used for antibody screening and serological crossmatching is now predominantly column agglutination with very few laboratories using tube techniques.

The currently available CAT systems include Ortho BioVue, which comprises a six-well cassette containing a glass microbead matrix, ID DiaMed cards utilizing Sephadex gel matrix, also with six wells, and an eight-well Sephadex gel matrix card from Grifols. Each offers a wide range of profiles and reagent systems.

The currently available solid-phase systems are Immucor Capture-R and Bio-Rad Solidscreen II. The Immucor system utilizes a range of red cell antigens adhered to the surface of a U-shaped microplate well. Sensitized red cells are then used as a marker. The Bio-Rad Solidscreen II is a solid-phase method for antibody screening and identification. The wells of the microplates are coated with protein A to allow the reaction to bind to the plate forming a solid-phase cell layer.

Individual laboratories need to make careful and informed decisions when selecting reagents for pre-transfusion testing. It is vital that any abbreviated testing in an automated or semiautomated system is carefully evaluated for the risks that could ensue if important controls were omitted when using this technology for blood grouping.¹¹

Laboratory information management systems (LIMS) store patient details and results of laboratory tests, allowing timely and accurate access to important information. In the transfusion department, IT systems have a much broader use and the updated BCSH guidelines for the specification and use of Information Technology (IT) systems in blood transfusion practice (2006) reflect this.¹² Where possible, using bidirectional or unidirectional interfaces to automated blood grouping analysers, IT systems are used to eliminate errors that can arise when a manual step is employed, including interpretation of test results.

Computer algorithms support the ‘electronic issue’ of blood to patients with a negative antibody screen without the need to perform an antiglobulin crossmatch and, in the UK in 2009, 46% of laboratories were using this system for some or all of their patients (Fig. 22.4). The use of automation for all aspects of compatibility testing is now recommended practice as it is recognized that it is safer.³ Automation brings several or all of the discrete activities of compatibility testing into a single platform process. It provides various levels of increased security over manual testing and may provide justification for abbreviated pre-transfusion testing (e.g. abandoning duplicate D, previously termed ‘RhD’, typing or reverse ABO grouping in the presence of a valid historical group). A risk assessment must be made and documented prior to any abbreviation of an established procedure, with consideration being given to the presence or absence of key functions in the automated equipment. The BCSH guidelines for compatibility procedures¹² and guidance from the MHRA¹³ give a list of factors to be taken into consideration and the reader is advised to consult these prior to implementing automated or semiautomated systems.

Figure 22.4 Change in proportion of UK laboratories issuing blood electronically 1998–2009. Data from UK NEQAS surveys indicating the increased use of electronic issue for compatibility testing in UK transfusion laboratories.

Pre-transfusion compatibility systems

The process of providing blood for transfusion involves many steps, all of which have to be reliably completed. These include the following:

• Blood samples have to be taken from the correct patient and labelled at the bedside in a single uninterrupted procedure. The sample must be identified by four core patient identifiers: the correctly spelled forename and surname, the exact date of birth and an accurate unique patient number (such as the NHS number or equivalent). The sample should be dated and signed by the person taking the sample.¹⁴ Some guidelines recommend that addressograph labels should not be accepted on blood samples for compatibility testing because it increases the risk of ‘wrong blood in tube’(WBIT).¹⁵ In an international study from the BEST collaborative, it was estimated that miscollected samples demonstrating WBIT occurred at a median rate of 0.5 per 1000, although there was great variation worldwide in the reported frequency of mislabelled samples, probably resulting from variation in policies for sample acceptance.¹⁶ Labels printed at the bedside using barcoded patient wristband and hand-held scanners are acceptable.¹⁴ The laboratory should have a policy for rejecting badly labelled samples, although in 2004 a survey of hospitals in England and North Wales showed considerable variation in the content and application of these policies.¹⁷

• A request for services should include the previously outlined four core patient identifiers as well as clear information about the source of the request and the location of the patient and clinical details including a detailed justification for the request. Previous transfusion history and any special considerations for the selection of blood (e.g. antigen negative if the patient has a clinically significant antibody, cytomegalovirus [CMV] tested or irradiated for certain groups of immunosuppressed patients) are also needed.

• An ABO and D group of the patient sample must be accurately performed.

• An antibody screen of the patient’s plasma (or mother’s plasma in the case of a neonate) should be able to detect any clinically significant red cell antibodies. In the event of a positive red cell antibody screen, antibody identification should be undertaken to assist the selection of compatible blood.

• There should be a check of existing transfusion records to compare current and historical results.

• The appropriate blood component should be selected and issued to a named patient using a serological crossmatch or electronic issue.

• Traceable documentation should exist to ensure that the results of laboratory compatibility procedures are available at the patient’s bedside to allow a check before transfusing the blood component. This should include a blood bag compatibility label (Fig. 22.5) and may include a compatibility form. The patient must be identified with a wristband and the blood component should be prescribed on a drug or fluid administration chart.¹⁴ In some countries, an additional bedside check of the patient’s blood group is undertaken prior to commencing the transfusion.

Figure 22.5 A unit of red cells showing the compatibility label.

Documentation of the Transfusion Process

All stages of the transfusion process must be clearly documented and these records must be kept. In addition to guidance from the UK Royal College of Pathologists on the retention of documents,¹⁸ the BSQR 2005 regulations stipulate that the records must be accessible for 30 years.⁷ This allows any blood component to be traced from the donor to the recipient should information come to light about any potential infective risks to the recipient.

Computer records are easier to search than paper records, but laboratory information systems are likely to become obsolete and be replaced several times within this mandatory 30-year period, so provision must be made to store historical data in an accessible format when procuring a replacement computer system.¹² Patient-held records are useful for patients who are treated in more than one institution, particularly if they have red cell antibodies and require phenotyped blood or if they have special requirements because of their underlying disease or its treatment. Credit card-sized records with corresponding patient information leaflets are issued by some transfusion centres to patients with red cell antibodies and similar cards exist for patients who require irradiated cellular blood components.¹⁹

Identification and Storage of Blood Samples

Depending on laboratory practice, blood transfusion tests use a clotted (serum) sample or EDTA-anticoagulated blood (plasma). Most laboratories using automated systems will use a plasma sample.

The reason for automated systems requiring plasma is because incompletely clotted blood samples may contain small fibrin clots that trap red cells into aggregates that could resemble agglutinates which could be falsely interpreted as a positive reaction. Clotted samples should be taken into a plain tube but not a tube with a separating gel.

The presence of complement in serum can cause lysis. If laboratory staff are used to recognizing agglutination as an indicator of a positive reaction they may fail to interpret the lysed red cells as an equally valid positive reaction. Therefore, when using serum for blood grouping and compatibility testing, any red cells in the test system should be washed and resuspended in saline which contains ethylenediamine tetra-acetic acid (EDTA) (see later). In addition, false-negative reactions may occur in immediate spin crossmatching with potent ABO antibodies, where rapid complement fixation causes a prozone effect (bound C1 inhibits agglutination). EDTA saline is not necessary when using plasma.

Throughout this chapter, for the sake of brevity, ‘plasma’ is used as the majority of UK laboratories are automated and therefore anticoagulated (plasma) samples are required. If serum samples are specifically required, this will be indicated in the text.

On being received in the laboratory, the details on the request form must be checked against the blood sample. Each blood sample must be labelled with a unique sample number. Barcode labels offer the advantage of positive sample identification and reduce the number of transcription errors. Samples inadequately or inaccurately labelled should NOT be used for pre-transfusion testing.¹¹

Great care must be taken to select and identify the sample prior to any testing. Transposition of samples in the laboratory can lead to an incorrect blood group being assigned to a patient, with serious consequences, including ABO incompatible transfusions. Where possible, the primary sample should be used for testing. If samples are separated, the plasma must be clearly and accurately identified and special precautions must be taken. If repeated testing on this sample is anticipated, storing separate small aliquots reduces the risk of sample deterioration, which occurs with repeated thawing/freezing of larger samples.

Whole blood samples should be tested as soon as possible because they will deteriorate over time. Problems associated with storage include lysis of the red cells, loss of complement in the serum and decrease in potency of antibodies. The BCSH guidelines ¹¹ indicate working limits as outlined in Table 22.1. If separated plasma or serum samples are stored for later serological crossmatch, care must be taken to ensure that the patient has not been transfused in the interim. It has been recommended that samples should be kept for a minimum of 7 days from group and screen, stored at 4°C. Samples should be retained post-transfusion for investigation of acute transfusion reactions and preferably stored for 7 days post-transfusion to enable investigation of delayed transfusion reaction.¹⁸

Table 22.1 Recommended limits for storage of samples used for compatibility testing ¹¹

ABO and D grouping

ABO and D grouping must be performed by a validated technique with appropriate controls. Before use, all new batches of grouping reagents should be checked for reliability by the techniques used in the laboratory. Grouping reagents should be stored according to manufacturer’s instructions.

ABO Grouping

ABO grouping is the single most important serological test performed in compatibility testing; consequently, it is imperative that the sensitivity and security of the test system are not compromised. The fact that anti-A and anti-B are naturally occurring antibodies allow the patient’s plasma to be tested against known A and B cells in a ‘reverse’ group.

This is an excellent built-in check for the ‘forward’ or cell group and has always been considered to be an integral part of ABO grouping, allowing the reading and recording of test results to be split into two discrete tasks. However, with secure, fully automated systems, linked to secure laboratory information management systems that, in combination, have the ability to prevent procedural ABO grouping errors, some laboratories now omit the reverse group when testing samples for which a historical group is available.¹¹ This should only be considered following a careful risk assessment and taking into account that the first sample taken may have been from the wrong patient. This may be in the order of 1:2000 samples.¹⁶ Any discrepancy between the forward and reverse groups should be investigated further and any repeat tests should be undertaken using cells taken from the original sample rather than from a prepared cell suspension.

Reagents for ABO Grouping

Monoclonal anti-A and anti-B reagents have replaced polyclonal reagents in routine grouping tests. A₁ and B cells are used for reverse grouping; group O cells or an auto control may be included to ensure that reactions with A and B cells are not a result of the presence of cold autoantibodies. A diluent control should be included where recommended by the manufacturer.

D Grouping

D grouping is usually undertaken at the same time as ABO grouping for convenience and to minimize clerical errors that may arise through repeated handling of patients’ samples. In the absence of secure automation, testing should be undertaken in duplicate because there is no counterpart of the ‘reverse’ grouping of ABO testing.

Reagents for D Grouping

Monoclonal reagents do not have the problem of possible contamination with antibodies of unwanted specificities, as was the case with polyclonal reagents. Therefore, the duplicate testing may be undertaken using the same anti-D reagent, although this should be dispensed as though it were two separate reagents.

DVI is the partial D with the fewest epitopes; therefore of all the D variants, DVI individuals are those most likely to form anti-D and a case of severe haemolytic disease of the fetus and newborn (HDFN) has been described.²⁰ For this reason, anti-D monoclonal reagents that do not detect DVI should be selected for testing patients’ samples.^21,²² The use of anti-CDE reagents has led to the misinterpretation of r′ and r″ cells in UK National External Quality Assessment Scheme (NEQAS) exercises and, because they are of no value in routine patient typing, their use is not recommended.^10,¹¹ Selection of high-avidity monoclonal anti-D reagents will allow detection of all but the weakest examples of weak D, negating the need to use more sensitive techniques to check the D status of apparent D negatives.

Some anti-D reagents have high levels of potentiator (e.g. polyethylene glycol) and should be used with caution; a diluent control is essential to demonstrate that the diluent does not promote agglutination of the test red cells as may happen if the patient’s cells are coated in vivo with immunoglobulin G (IgG); any positive reaction seen with the control, however weak, invalidates the test result. Anti-D reagents are provided in many different kits and those responsible for selection and purchase should make themselves aware of the content, specificity and potentiation of the chosen reagent.

Column agglutination techniques

CAT techniques (see Chapter 21, p. 502) are now the commonest method for grouping in the UK (80% of laboratories who responded to a UK NEQAS survey in 2009, see Fig. 22.2), especially where automated systems are in place; these should always be performed in accordance with the manufacturer’s instructions. There are several different profiles to choose from and some cards/cassettes include monoclonal antibodies to other blood group antigens (e.g. K) in addition (see Chapter 21, p. 502 and Fig. 22.6). Forward and reverse grouping may be undertaken in separate cards/cassettes.

Figure 22.6 Column agglutination technology: blood group O D positive. First box, α. Second box, β. Third box, D. Fourth box, Control. Fifth box, A₁. Sixth box, B.

Solid-phase techniques

ABO/D grouping using solid-phase techniques would usually be part of a fully automated system and testing should be accordance with the manufacturer’s instructions. Bio-Rad Erytype employs a standard agglutination technique for blood grouping performed in a microtitre plate. The reagents for the forward group are dried onto the wells. Uncoated empty wells are used with standard liquid red cells for a reverse group.

Controls

Positive and negative controls should be included with every test or batch of manual tests. In fully automated systems, the controls should be set up at least twice in a 24-h period. The timings should coincide with machine startup and changing of reagents, taking account of the length of time that reagents have been kept at room temperature on the machine. The control samples should be loaded in the same way as the test samples. The required controls are shown in Table 22.2. Where controls do not give the expected reactions, investigations should be undertaken to determine the validity of all tests undertaken subsequent to the most recent valid control results.

Table 22.2 Control cells for blood grouping

Reagent	Positive control	Negative control
Anti-A	A cells	B cells
Anti-B	B cells	A cells
Anti-D	D positive cells	D negative cells
A₁ cells	Anti-A	Anti-B
B cells	Anti-B	Anti-A

Causes of Discrepancies in ABO/D Grouping

False-Positive Reactions

Rouleaux

Rouleaux may occur in various clinical conditions, where the ratio of normal albumin to globulin is altered in plasma (e.g. multiple myeloma) and in the presence of plasma substitutes such as dextrans. The stacking of red cells on top of one another in columns may be misinterpreted as weak agglutination by inexperienced workers. Rouleaux will usually disperse on a slide if a drop of saline is added; alternatively, the reverse grouping can be repeated using plasma diluted 1 in 2 or 1 in 4 with saline.

Cold autoagglutination and cold reacting alloantibodies

Cold autoantibodies (usually anti-I) that are reactive at room temperature may lead to autoagglutination in the cell group (ABO and D) and panagglutination in the reverse group, causing a grouping anomaly (p. 277). The tests should be repeated using cells washed in warm saline and plasma pre-warmed to 37°C. An auto control should be included.

Cold reacting alloantibodies (e.g. anti-P₁) may cause agglutination of reverse grouping cells and, if this is suspected, reverse grouping should be repeated using pre-warmed plasma or grouping cells that lack the implicated antigen.

T-activation/polyagglutination

Polyagglutination ²⁵ describes agglutination of red cells by all or most normal adult sera but not by the patient’s own serum. This is as the result of IgM antibodies reacting with an antigen on the red cells which is usually hidden but can be exposed by enzyme activity. The most common form is T-activation, which occurs when the bacterial enzyme neuraminidase cleaves N-acetyl neuraminic acid from the red cell membrane, exposing the T antigen.

This used to be a problem when grouping with polyclonal reagents, which contain anti-T, but tests using monoclonal reagents are not affected by this phenomenon.

Acquired B

This arises in group A₁ patients where the expected reaction with anti-A is noted but an additional weak reaction with anti-B occurs. The blood group appears to be ‘AB’ but with anti-B in the reverse group. This discrepancy between the forward and reverse grouping may be overlooked if the patient’s own anti-B is weak or if the reverse group is omitted.

The acquired B antigen is usually caused by a bacterial deacetylase enzyme acting on A₁ red cells and producing a B-like substance. Some anti-B reagents react strongly with the acquired B antigen (e.g. those derived from the ES4 clone).²⁶ Such anti-B reagents are rare but should be avoided in routine blood grouping.

Potentiators

Red cells may be coated with IgG as a result of in vivo sensitization. The use of potentiated techniques, such as the antiglobulin test for D typing, or of potentiated reagents for ABO or D typing, may result in a false-positive reaction; the latter would also result in a positive reaction with the diluent control, but UK NEQAS data have shown that some laboratories fail to include an appropriate control or fail to understand the significance of a positive control.²⁷ For this reason, use of potentiated techniques or reagents for blood grouping is not advised.

The most likely cause of a false-negative grouping result is failure to add the reagent or test plasma. For this reason, in liquid-phase tube and plate tests, the serum or plasma should always be added first to the reaction chamber and a visual check should be made before red cells are added. The use of colour-coded reagents for ABO grouping is helpful in this respect. Automated systems have liquid level sensors resulting in an alert of the failure to add a reagent or test plasma.

Loss of potency

Inappropriate storage or freezing and thawing may cause a loss of potency of blood grouping reagents. The use of regular controls will alert the user to this problem.

Failure to identify lysis

In the presence of complement, anti-A and anti-B may cause in vitro lysis of reagent red cells. If lysis is not recognized as a positive reaction, falsely negative results may be recorded in reverse grouping tests. To avoid this, reverse grouping cells should be resuspended in EDTA saline, where serum rather than plasma is used.

Mixed-field appearance

This describes a dual population of agglutinated and non-agglutinated red cells which may be observed in both ABO and D grouping. It is important to recognize this as a mixed-field picture and not to confuse it with weak agglutination. The most likely cause of a mixed-field picture is the transfusion (either deliberate or accidental) of non-identical ABO or D red cells. Investigation will be required to determine the actual blood group of the patient, who may have been transfused in an emergency, at a different establishment or may have received an intrauterine transfusion. A mixed-field ABO group may be the first indication of a previous ABO-incompatible transfusion.

An ABO or D incompatible haemopoietic progenitor cell transplant will result in a mixed-field picture until total engraftment has occurred; the mixed-field picture may subsequently reappear when a graft is failing. Rarely, a dual population of cells is permanent and results from a weak subgroup of A (A₃) or a blood group chimerism.

Interpretation of a dual population of red cells will depend on the technique used. In a tube, microscopic reading will reveal strong agglutinates in a background sea of free cells. In column agglutination techniques there will be a line of agglutinated cells at the top of the column, with the non-agglutinated cells travelling through to the bottom of the column. In liquid-phase techniques if the reaction grade is not a strong positive or an obvious negative then the reactions requires further investigation, which may include microscopic examination. In a solid-phase technique a mixed field is seen as a dual population of cells, with the agglutination being surrounded by free cells. Automated systems should be set up to detect mixed-field pictures and this should be used in conjunction with local policies.

D variant phenotypes

Weak D phenotypes are where the entire D antigen is present but there are fewer D antigen sites per cell and most weak D types group as D positive with the currently available high-avidity commercial monoclonal anti-D reagents. Where differing reactions are obtained with two reagents, the patient may be a partial D (i.e. one or more of the epitopes of the D antigen is missing). It was generally thought that patients who are weak D are unable to make anti-D and may be treated as D positive whereas some patients with partial D may be capable of making immune anti-D following sensitization with the missing epitope. This concept has been challenged by reports of weak D patients with anti-D.²⁸

Detailed genotypic analysis combined with structural modelling of the D antigen in D variants have demonstrated change in the amino acid sequence in the extracellular domain in partial D, whereas weak D is associated with mutations resulting in change to the amino acid sequence in the intracellular or membrane-spanning domain.

A weak reaction with a single anti-D reagent should be investigated with a second anti-D reagent before assigning a result of D positive. It is essential to be able to distinguish between a weak reaction and a mixed-field reaction because the latter may be the result of a patient who is being transfused with blood of a different D type.

If in doubt, it is safer to call the patient D negative, at least until investigations have been undertaken by a reference centre. This will be of no clinical consequence because it is safe to transfuse D negative blood to a patient who is D positive, and a pregnant woman who is D positive and her unborn child would be unlikely to be harmed by the injection of prophylactic anti-D immunoglobulin.

Current advice about choice of D-typing reagents is that because DVI lacks the most epitopes, such individuals are likely to make anti-D when challenged by transfusion or pregnancy. For this reason, anti-D reagents for routine grouping of patients’ samples should not detect DVI.¹¹ There is little evidence to suggest that a DVI donor would elicit an immune response in a recipient who is D negative; however, weak D positive and partial D donors, including DVI donors, should be classified as D positive.²⁹ There are some important ethnic differences in the frequency of different Rh haplotypes, as shown in Table 22.3. Resolution of anomalous D grouping where a partial D is suspected now includes both serological testing and genotypic studies.³⁰

Table 22.3 Approximate frequencies of common haplotypes in selected populations ³⁰

Antibody screening

Antibody screening is usually undertaken at the same time as blood grouping and in advance of selecting blood for transfusion. Antibody screening may be more reliable and sensitive than crossmatching against donor cells because some antibodies react more strongly with red cells with homozygous expression (double dose) of the relevant antigen than with those with heterozygous expression (single dose) – most notably anti-Jk^a/Jk^b but also anti-Fy^a, -Fy^b, -S and -s. Screening cells can be selected to reflect this, whereas donor cells are usually of unknown zygosity. In addition, reagent red cells are easier to standardize than donor cells and there is potentially less opportunity for procedural error, particularly in automated systems.

Clinically significant antibodies are those that are capable of causing patient morbidity as a result of accelerated destruction of a significant proportion of transfused red cells. With few exceptions, clinically significant antibodies are those that are reactive in the indirect antiglobulin test at 37°C; however, it is not possible to predict serologically which of these antibodies will definitely be of clinical significance, so the term ‘of potential clinical significance’ is often used.

Red Cell Reagents

The patient’s plasma should be tested against at least two individual screening cells, used individually, not pooled. The screening cells should be group O and encompass the common antigens of the indigenous population.

In the UK, the following antigens should be expressed as a minimum: C, c, D, E, e, K, k, Fy^a, Fy^b, Jk^a, Jk^b, S, s, M, N and Le^a; one cell should be R₂R₂ and another R₁R₁ or R₁^wR₁. The following phenotypes should also be represented in the screening set: Jk(a+b−), Jk(a−b+), S+s−, S−s+, Fy(a+b−) and Fy(a−b+) (Table 22.4). These recommendations for homozygosity are based on UK data regarding the incidence of delayed haemolytic transfusion reactions, the need for high sensitivity in the detection of Kidd antibodies and the poorer performance of column agglutination techniques in the detection of some examples of Kidd antibodies using heterozygous cells.^11,²⁹ The requirement for the expression of C^w and Kp^a antigens on screening cells has been the cause of much debate but in the UK and the USA detection of anti-C^w or anti-Kp^a is not a requirement even in the absence of an antiglobulin crossmatch.^11,²⁹ This is because these are low-frequency antigens and the antibodies rarely cause delayed haemolytic transfusion reactions or severe haemolytic disease of the newborn.